AI Pavement Inspection

Potholes and cracks on roads affect road safety. In our role as providers of road inspection and maintenance projects in various districts of Taipei City, it is our responsibility to protect the roads. In order to make inspections more efficient, we have replaced road-users as the primary method of identifying of road safety issues. The company introduced the AI identification system in 2019, we are leaders in the industry in introducing smart technology into road inspection operations, replacing the traditional human eye to judge road defects, making road inspections and repairs fast, convenient, and accurate.

​

The company developed a panoramic inspection vehicle and set up a high-resolution panoramic camera and image recognition system. When the vehicle is driving on the road, it can collect 360-degree images of the road through the panoramic camera. The image is sent back through the system to automatically generate a 3D model, which presents the true road and street model. Coupled with the pavement management system and interpretation and processing through artificial intelligence (AI), the model can accurately detect data such as pavement damage, smoothness, and rutting depth. Based on the performance indicators, it can automatically identify road defects and discover which roads require maintenance. We have applied this technology to our current road maintenance projects to make inspections faster and more accurate, greatly reduce the burden of manual inspections, and indeed improve operational efficiency.

​

The company researches and develops the surrounding view detection vehicle, and sets up a high-resolution surrounding view camera and an image recognition system. When the vehicle is driving on the road, it can collect 360-degree road images through the surrounding view camera, and send the image back through the system to automatically generate a 3D pattern, which is realistic. Presenting road and street view models, combined with pavement management system and artificial intelligence interpretation processing, can accurately detect data such as road pavement damage, flatness, and rut depth, and automatically identify road defects based on detection and maintenance indicators, and find out which roads need Carry out maintenance processing, and we have applied this technology to current road maintenance projects to make inspections faster and more accurate, and greatly reduce the burden of manual inspections, and indeed improve operational efficiency.